ST2000XM and 9x9 Binning?

Discussion in 'Legacy Models - Community Support' started by f11, Feb 28, 2017.

  1. f11

    f11 Cyanogen Customer

    Joined:
    Apr 9, 2015
    Messages:
    8
    Location:
    Edmonton, AB Canada
    I use an ST2000XM camera with CFW8 filter wheel and an AO-8. Its being controlled by TheSkyX Pro with Camera Add-On from a 2009 vintage iMac running Mavericks 10.9.5 (stayed here due to USB problems in newer OS releases). The camera has been initiated with the most recent universal driver for OSX from SBIG.

    The camera launches and does everything expected of it, except when doing a focus run using @Focus2: the first focus image in a focus run looks like a VERY zoomed in picture, and contains FITS Header entries XBINNING and YBINNING set to 9 ... that is, the image is considered to be a 9x9 binned image. Its always 177 x 133 pixels in size, which is almost exactly 1/9th the size of my main imaging chip at 1600 x 1200.

    But the ST2000XM specs show its only capable of 3x3 binning, and when I do my focusing I've set binning to 1x1.

    Going back in my records, I se that on Feb 12th 2013 and before, my focus runs (TSX Pro w/ C-AO under Windows back then) always started with the expected 1x1 image. The next night, Feb 13th, and ever after my focus runs always contain the mysterious 9x9 first image. I was having some focus issues a couple nights ago, and noticed this anomaly, and traced it back to the Feb 2013 imaging runs. My images also contain a FITS Header entry for the controlling software, so I know I didn't update my TSX software between the two nights.

    I appreciate this isn't the SB TSX forums, so not looking for help with that program. But since the 9x9 data is recorded in the FITS Header of the images, which I THINK is put there by the camera firmware/software, and not by TSX, has anyone ever seen anything like this? Or was there some fundamental change in the firmware or software in Feb 2013 that would account for this? I can't find a FITS Header entry for the camera's FW or SW versions, and can't recall upgrading the camera between those two nights.

    The focus runs complete normally, except for this extra bizarre image, so its more of an annoyance than anything so far. But I worry its a symptom of something else that might not be so benign.

    Any ideas here? Grateful for any light that can be shed...
    Rod
     
  2. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    9,952
    The software builds the FITS header, not the driver. The driver doesn't know anything about FITS.

    It is not unusual for software to implement binning above and beyond what the camera can actually do.
     
  3. f11

    f11 Cyanogen Customer

    Joined:
    Apr 9, 2015
    Messages:
    8
    Location:
    Edmonton, AB Canada
    Thanks, Doug - that really helps !! The camera is (literally) a black box to me, so I often don't know where some processes originate, camera or TSX.

    So, regardless of the camera's ability to do so, software could request it to take an image at 9x9. The camera wouldn't balk at an out of spec request? What would the result of such a request look like for the st2000xm? And the camera doesn't know anything about the FITS table entries and doesn't write to that data table. So the problem lies squarely in TSX @Focus2 functions, not in the camera. That's good!

    What command that the camera understands would be required to request a 9x9 image from a st2000xm? I'd like to see if I can capture the request as a troubleshooting clue for when I go back to the TSX forums.
     
  4. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    9,952
    The software could tell the camera to do a 3x3 bin, and then software bin the resulting image to 9x9.
     
  5. f11

    f11 Cyanogen Customer

    Joined:
    Apr 9, 2015
    Messages:
    8
    Location:
    Edmonton, AB Canada
    Can you think of any conceptual reason a focus function might want to do this, even just to initialize? So far, nobody else on TSX forums has reported seeing this behaviour, which points the finger at me (configuration errors and so on). But for the life of me I have no idea what I might have done, or more specifically why.
     
  6. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    9,952
    Sorry I really can't help with third party software. I don't know anything about how it is built. I can only answer basic questions about what is possible, etc.
     
  7. f11

    f11 Cyanogen Customer

    Joined:
    Apr 9, 2015
    Messages:
    8
    Location:
    Edmonton, AB Canada
    I understand that, Doug and wouldn't ask you to guess at what TSX is doing - how are you supposed to know? :)

    But I was asking a more generic question: is there any practical reason you can think of for requesting a 9x9 full-chip image as the initial step of a focusing process where the user-set binning is 1x1 with a sub-frame of 200x200? Is there some rationale that you're aware of that might explain it?

    It just seems to me that the initial image should be "at current focus" with binning and sub-frame per user-request. For example, for MaximDL's focus function ... what is the first image taken?
     
  8. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    9,952
    Sometimes software will use binning during the "find star" process to speed things up. 9x9 is extreme though and unlikely to work well.

    MaxIm DL only does exactly what you tell it to.
     
    f11 likes this.
  9. f11

    f11 Cyanogen Customer

    Joined:
    Apr 9, 2015
    Messages:
    8
    Location:
    Edmonton, AB Canada
    Ok, I think that's it for my questions for now. Thanks for y0ur patience, Doug ... you've helped me narrow things down a bit.

    BTW, I like software that does only what its told to do. :)
     
  10. f11

    f11 Cyanogen Customer

    Joined:
    Apr 9, 2015
    Messages:
    8
    Location:
    Edmonton, AB Canada
    As closure on this thread - I just got confirmation from SB that if I set up @Focus2 to automatically place a subframe around the target star, TSX attempts to optimize the size of the initial full-frame image to minimize download time, by binning as aggressively as possible. So we know why its happening, but still waiting to hear why I'm getting a 9x9 bin instead of the camera's maximum spec of 3x3. It might be because I choose the SBIG with AO camera: as a result, TSX may not be able to determine which model my camera is and just assumes the 9x9 mode.

    Thanks again for your time, Doug!!
     

Share This Page